The algorithm has retrieved 134 women, for a total of 240 pregnancies, from 2000 to 2020.
543 records were reconciled to build the pregnancies.




Streams


Pregnancies are retrieved from four different streams:

Page Title
  • PROMPTS: records from birth records, pregnancy termination records, and abortion records in SURVEY_ID: the existence of one of these records easily implies that a pregnancy is terminated

  • EUROCAT: records from EUROCAT table

  • CONCEPTSETS: diagnostic codes from EVENTS or procedure codes from PROCEDURES or codes from the MEDICAL_RECORDS file that refer to a terminated or ongoing pregnancy

  • ITEMSETS: routine health variables that are populated only when a woman is pregnant


A pregnancy may have records in different streams. The subsequent figure shows the number of absolute pregnancies, divided by the highest quality stream in with the pregnancy is retrieved, and anchored to the pregnancy start year.






Record Quality


As reported in the sap, some records have both pregnancy start date and pregnancy end date available, for others one or both of these dates are imputed. Therefore, the quality of the record depends on both the stream and the imputation. The complete list with the hierarchy of the fifteen different qualities is given below:


Page Title
  1. EUROCAT: both pregnancy_start_date and pregnancy_end_date are recorded

  2. PROMPT: both pregnancy_start_date and pregnancy_end_date are recorded

  3. ITEMSETS: both pregnancy_start_date and pregnancy_end_date are recorded

  4. CONCEPSETS: pregnancy completed and pregnancy_start_date recorded

  5. PROMPT: pregnancy completed and pregnancy_start_date not available and imputed

  6. ITEMSETS: pregnancy completed and pregnancy_start_date not available and imputed

  7. CONCEPSETS: live birth, meaning non primary care, pregnancy_start_date not available and imputed

  8. CONCEPSETS: pre-term birth, meaning non primary care, pregnancy_start_date not available and imputed

  9. CONCEPSETS: still birth, meaning non primary care, pregnancy_start_date not available and imputed

  10. CONCEPSETS: interruption, meaning non primary care, pregnancy_start_date not available and imputed

  11. CONCEPTSETS: spontaneous abortion, meaning non primary care, pregnancy_start_date not available and imputed

  12. CONCEPTSETS: ectopic pregnancy, meaning non primary care, pregnancy_start_date not available and imputed

  13. CONCEPTSETS: meaning implying primary care, pregnancy_start_date not available and imputed, end date estimated with record date

  14. all Streams: ongoing pregnancy and pregnancy_start_date recorded

  15. all Streams: ongoing pregnancy having pregnancy_start_date not available and imputed


The subsequent figure shows the number of absolute pregnancies retrieved, divided by the highest record quality in the pregnancy, and anchored to the pregnancy start year.

The number of pregnancies is also shown in the table below.






Type of pregnancies end


The same analysis is done for each of the five different types of end of pregnancy, which are:

  • LB : live birth

  • SB : still birth

  • T : termination

  • SA : spontaneous abortion

  • ECT : ectopic pregnancy

  • UNK : unknown



Live Birth




Still Birth


## [1] "there is no SB among the types of end of pregnancy"



Termination




Spontaneous Abortion




Ectopic




Unknown







First record



For each pregnancy the difference (in days) between the date of the first record retrieved and the date of beginning of pregnancy is calculated. The sebsequent graphs show the distribution of this distance for each type of end of pregnancy.



Live Birth




Still Birth


## [1] "there is no SB among the types of end of pregnancy"



Termination




Spontaneous Abortion




Ectopic




Unknown







Excluded Pregnancies



The graph shows the absolute number of excluded pregnancies for each year, divided by reason for exclusion.





The total number of excluded pregnancies is 40, the reason is shown in the following table.








Reconciliation


In the reconciliation outcomes, two records of the same pregnancy can be:

  • concordant: equal start and end dates;

  • slightly discordant: start and end date different by no more than 7 days;

  • discordant: start and end date different more than 7 days;

  • inconsistent: when a red record is not included between the beginning and the end of the pregnancy.

The next graph shows the percentage of pregnancies that have non-concordant records. A pregnancy can be composed of both concordant and discordant records, the graph shows the number of pregnancies with at least one of the outcomes.






Dummy Tables




Exclusion



In the next table the absolute number of pregnancies included and excluded is reported.


Pregnancies N (%)


The following table shows the reason for exclusion:

  • GG_discordant_end : two green records are discrepant on the date of end of pregnancy

  • GG_discordant_start : two green records are discrepant on the date of start of pregnancy

  • Insuff_quality: there are no records of sufficient quality in the pregnancy


Pregnancies N


The next analysis will take into account only those pregnancies that have not been excluded



Age



The following table describes the distribution of the age of start of pregnancy, divided by type of end of pregnancy.



The next table describes the absolute number of pregnancies (and the percentage in parentheses) divided by age bands and type of pregnancy end.

Pregnancies N (%)



Record



The following table shows the mean of records retrieved for each pregnancy. In parentheses is the standard deviation.


Mean (SD)


For each pregnancy multiple records are reconciled, the next table describes what quality of records represents the pregnancies.


Pregnancies N (%)



Meaning



The following table describes the meaning end dates.


Pregnancies N (%)


The following table describes the meaning start dates


Pregnancies N (%)